[harrison]

Hello. Please read the supplementary information of the paper here: https://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.3000794
It explains it in detail. Both how the data is produced, what the Emit/Excite/Base bands mean, and also what the baseline reading is that should be subtracted.

[harrison]

Hello, did you do the things above – e.g. using non-fluorescent media and making sure the cells are FAIRLY fluorescent? I.e. not very low copy fluorescent reporters…

[harrison]

Hello,
I think it is a failure in the USB connection between the Beaglebone and the host PC. One thing you might try is putting independent power into the beasglebone (it has a port for this); if the beaglebone is doing a lot of calculations it MIGHT draw too much power over USB and crash, so supplying independent power would fix this. Try that first, as well as getting a new USB cord (as short as possible) to link two.

[harrison]

Hello Giacomo
I have no idea, I am not familiar with this approach. You could integrate it with the current setup by using the inbuilt pause command, then run your python (which only needs to send I2C commands, though likely you should adopt the I2ccom subsystem from the Chi.Bio OS since it needs to go through a multiplexer).

[harrison]

I wouldnt have thought so. If it is getting _worse_ over time it certainly seems unlikely to be the culprit.

[harrison]

I am not sure I follow – i.e. are you proposing that low temperature will help the digital communications?

If it is just doing occasional errors I would leave it as is. If you think it is getting worse then I would consider replacing the microUSB cables as you may have bad ones.

[harrison]

The LED is pretty high power. Likely significantly more powerful than most external light sources. The real problem is filter bleed-through i.e. the amount of excitaiton light that makes it through the emission filter. Thus if you want better sensitivity best approach is to use an off-the-shelf spectrometer (i.e. something much better than the AS7341 filtered photodiode array device in there) as this will give much better spectral resolution. Does this make sense?
If you do want to try that, there are some good ones from Ocean Optics for ~1k mark that are easy to use and good price vs performance

[harrison]

Hello giacomo,
That setup sounds sensible.

How bright would you say it is as a function of “the brightest GFP cells you have ever seen”. If it is <0.5% of “very bright” it will be difficult to measure due to other autofluorescence and filter bleed-through. This consideration will depend on what the plasmid copy number/RBS/Promoter strengths all look like.
Also, what OD level are you regulating it at? Something like 0.5 is probably optimal, if you have it very low (e.g. 0.05) then the signal-to-noise may be more tricky as well.

[harrison]

Hi Mpatterson,

It is an error arising during the I2C digital communications, often due to (for example) the signal from the reactor to the pump having a lot of noise in it potentially leading to issues. If you have already cleaned the moisture sensors and cable ports there isn’t a whole lot you can do with it. Shorter micro-USB cables would help, as would taking the reactors away from any kind of massive electrical noise source if there was one nearby, though probably this is unlikely.
If it is still running happily enough I suppose let it go as-is. If it gets worse maybe try sourcing some different micro-usb cables and swap them out, perhaps the ones you have are from a dodgy manufacturing batch…

Harrison

[harrison]

Hello,

Are you sure you have internet connection to the computer you are running ubuntu on? It needs to have active internet to work.
Also, does 192.168.7.2:3000 work? (That should be the cloud9 code editor, if it doesn’t work then the problem is likely not the internet)

[harrison]

The “winch” signal is something to do with the GUI/server presenting data; generally not an important one. If it is giving issues after restart I’d recommend power cycling everything.
In general strongly recommend disabling Windows update or any similar auto-restart functionality on the host computer…

[harrison]

Potentially outlet pump may not be working effectively, i.e. if the tubing is not settled in the pump rotor well, or if it is very old tubing and has been ground down over time and hence makes a poor seal. I’d try pulling the tube through the pump head (so there is a fresh bit inside the pump) and see how it performs.

[harrison]

It looks as if the digital communication on the device may have falled mid-way through it trying to read data from the thermometer. Not sure that there is a particularly convenient fix or reason for this; was there any liquid on the top of the reactor afterward?

[harrison]

Have just emailed you with .step files 🙂

[harrison]

Hi Niko,

Usually if it is regulating at 0.4 OD it should be something like 0.36 to 0.44. Some due to variability in measurements and some due to actual variation in the OD due to pumping in and out.
Indeed taking the reactor in and out of the device will change the geometry slightly and this will change the OD. If you look in the operation manual it talks about how to get OD measurments as reliable as possible – in short start the experiment with blank media (no cells), then let it arrive at correct temperature, then zero OD, then add cells through lid. This will typically reduce variability by more than 50%.

For sampling, IMO the best approach is a long GC needle or other blunt needle that you can put through the lid WITHOUT taking the tube out of the device, and thus maintain alignment as above.

If you want diffusers please email me your mailing address and I can send a few.

[Author]

Posts